JITTER

To check for robustness of the proposed models, we have performed several additional tests. Taking in consideration the nested nature of our data, we added two additional models to account for possible unobserved differences in employees’ scores of innovative behaviours across the five business units in Energetica. Such differences might generate clusters in the error structure, thus affecting standard errors and significance levels. To account for these potentially unobservable differences that may affect the interpretability of the results, we augmented Model 3 by specifying cluster-robust standard errors at business-unit-level (White, 1984). By looking at Model 4 (see Table 4 above), we can see that in presence of the interaction term, the effect of brokerage became insignificant in the model with the outliers (b = 8.39, p > 0.10), but remained highly significant in the model that excludes the outliers (b = 10.97, p < 0.01). What is particularly important is that the hypothesized joint effect between brokerage and network diversity remains significant at 0.05 level in the model with the outliers when we specified cluster-robust standard errors at business-unit-level (b = -21.12, p < 0.05). When the outliers were excluded, the joint effect remained significant at 0.01 level (b = -26.36, p < 0.01). Taking in consideration that our dependent variable is based on supervisor ratings, in Model 5 we specified cluster-robust standard errors at supervisor-level. When considering a sample with the outliers (see the left-hand side part of the column under Model 5), the hypothesized interaction effect became weaker but still significant at 0.10 level (b = -21.12, p < 0.10). However, in a model without the outliers, the joint effect remains highly significant and considerably stronger than in a model with the outliers (b = -26.36, p < 0.01).

As an additional robustness check for supporting the proposed interaction effect, we augmented Model 3 by introducing a variable that controls for supervisor’s span of control

152 (Carnabuci and Diószegi, 2015). We operationalized supervisor’s span of control as a ‘number of subordinates’ that directly report to the supervisor, whereas the employees without supervisory role have been assigned a value of zero. Including this variable did not substantially affected our estimates of interest. The regression coefficient of the interaction effect slightly decreased but remained significant at 0.05 level (b = -20.66, p < 0.05).

Hence, the outcomes of all the additional checks that we have performed to further support the hypothesis 2, suggest that our results are robust and that the hypothesized moderation role of network diversity in the relationship between brokerage and innovative behaviour is negative. The theoretical and practical implications of these results are discussed in the next section.

DISCUSSION

In this paper, we studied the conditions under which social network brokerage affects innovative behaviour. We found that network diversity with respect to professional background negatively moderates this relationship. This finding has three specific, and equally important contributions to the extant literature. First, by showing that network diversity moderates the relationship between brokerage and innovative behaviour, the present study contributes the contingency view of social networks (Carnabuci and Diószegi, 2015; Tortoriello and Krackhardt, 2010). Second, the present study has important implications for the literature that study the effects of social structures on innovative behaviour (Perry-Smith and Mannucci, 2015). And third, we contribute to the diversity theory and research (Gilson et al., 2013; van Knippenberg et al., 2004; 2011), which has traditionally focused on diversity within formally defined organizational units, while we shed light on diversity within the informal networks of advice relationships. In the

153 following sections, we discuss the two theoretical contributions, its practical implications, and the apparent limitations, which make room for further research into the field.

Theoretical contributions

The support for our baseline hypothesis (H1) does not offer per sè, particularly novel insights. It empirically confirms that one of the most widely used structural variables in the social network literature, that is brokerage, is positively related to a desired organizational outcome such as innovative behaviour. However, the empirical support that we have found for the baseline hypothesis indicates that brokerage is indeed, an important, and easily generalizable construct that has strong predicting power on the variance of individuals’ innovative behaviour. This certainly justifies our efforts to further investigate the moderators that better explain this relationship. By showing that network diversity with respect to professional background works in conjunction with brokerage to predict innovative behaviour, we contribute to the recent debate in the extant literature about the conditions under which differences in social structure affect innovative behaviour (e.g. Carnabuci and Diószegi, 2015; Tortoriello and Krackhardt, 2010).

The empirical support for our H2, where we theorized a negative joint effect between brokerage and network diversity on innovative behaviour sheds new light on the contingency view of social networks. Focusing on the role of individual cognition, the past research has shown that the relationship between brokerage and innovative behaviour is contingent upon several personal characteristics such as self-monitoring personality (Mehra et al., 2001), and cognitive style (Carnabuci and Diószegi, 2015). These studies have advanced our understanding on how the same social structure can lead to differences in the innovative behaviour of brokers depending on their personal characteristics. However, the environmental conditions that enable or impede the brokers

154 to integrate and process the information and knowledge that come from distinct groups of a network, remained under-investigated (Tortoriello and Krackhardt, 2010). Although the exposure to diverse information and knowledge is critical for brokers to successfully generate novel and useful ideas (Ahuja, 2000), it also imposes challenges for brokers to integrate and make sense of such diverse information and knowledge (Ahuja, 2000). By focusing on professional background, as a category that depicts the knowledge/informational diversity among employees in an organizational setting, this paper shows that the relationship between brokerage and innovative behaviour is contingent upon network diversity with respect to the professional background of advice providers. We argue that brokers will enhance their innovative behaviour if they seek advice (i.e., information and knowledge) from advice providers that have the same professional background as the brokers have. Seeking advice from colleagues having the same professional background will stimulate a development of shared narratives and meanings about the surrounding concepts (Boisot, 1995), thus facilitating communication, and enhancing the ability of advice seeker (i.e., brokers) to successfully integrate the provided information and knowledge (Dougherty, 1992). Indeed, as past research has argued, it is easier to integrate the available information and knowledge if the parties that are involved in such an exchange share some common knowledge base (Ahuja, 2000). Thus, we contribute to the contingency perspectives of social networks by showing that beyond the personal characteristics (i.e., self-monitoring, cognitive style), the environmental characteristics (i.e., the professional background of the advice providers) as well play an important contingent role in the relationship between brokerage and innovative behaviour.

The present study contributes also to the diversity literature, which traditionally has focused on diversity within formally demarcated organizational units such as groups, teams, and

155 organizational departments (Cheung et al., 2016; van Knippenberg et al., 2010). This line of inquiry has for long studied the effects of diversity within formal units on that units’ performance (Cheung et al., 2016; van Knippenberg et al., 2010). The scholars have only recently started to explore diversity from a multilevel perspective, and it has been shown that group diversity is conducive to individual-level creativity (e.g. Gilson et al., 2013). The present study makes a step forward by highlighting the importance of diversity in informal relationships where individuals engage because of instrumental motives. Towards this end, we approached the concept of diversity from a social network perspective. To the best of our knowledge, the concept of network diversity has been applied as a feature of social networks that can affect desired organizational outcomes but not as a contingency variable (Baer et al., 2015; Chen and Gable, 2013). For instance, the study by Chen and Gable (2013) has conceptualized network diversity as the degree to which an ego has formed ties with alters belonging to different cohorts as well as to the relative distribution of ties among the cohorts. The cohorts in this study are defined in terms of belongingness of alters into a corporate department and belongingness of alters to higher hierarchical positions. Hence, the categories over which the network diversity is calculated mirrors the most important features of a formal organization such as structured communication flows (i.e. departments) and legitimate authority (i.e. higher hierarchical positions) (McEvily et al., 2013). Instead, in the present study we focus on professional background as a category over which we operationalized the network diversity. The professional background is not a formally defined organizational unit but still can capture the informational/functional diversity (Anteby et al., 2016). Moreover, given the divisional structure of the company individuals belonging to the same professional background could work in different organizational business units (e.g. economists work at the Sales department of the five business units), and thus, the studies that operationalize diversity with respect to formally defined

156 categories such as department, may overestimate the effects of diversity on desired organizational outcomes. Our objective was to understand how the concept of diversity with respect to professional background, as a category that captures the informational/functional diversity, can affect individual-level outcomes such as innovative behaviour. By showing that professional background diversity works in conjunction with brokerage to predict innovative behaviour, we offer an additional moderator that can be integrated within the CEM model proposed by van Knippenberg et al. (2004), which posits that diversity effects on desired organizational outcomes are better predicted if studied in conjunction with a range of possible moderating variables.

Practical implications

Our study offers important insights for the managers and practitioners, who want to enhance the innovative behaviour of their employees. The empirical support that we found for the baseline hypothesis suggest that apart and beyond the formal organization, managers and practitioners should focus on the patterns of informal relationships among the employees. Our findings show that social capital that stem from closed as well as brokering network positions can be beneficial for innovative behaviour. Collecting social network data thus, can be a useful practice aimed at increasing the social capital among the employees (Cross and Parker, 2004). By drawing the informal map of relationship at the workplace, managers and practitioners can understand the patterns of information and knowledge flow among the employees. Since the baseline hypothesis predicted that brokering network position will be positively related to innovative behaviour, managers and practitioners should ensure that the informal network relationships is rich in structural holes. This can be achieved by stimulating the informal interaction among seemingly disparate organizational units. This can be done by bringing the members from distant organizational units together at formal as well as informal organizational meetings or events.

157 The second practical contribution of this study stems from the support that we found for our second hypothesis, which suggests that brokerage interacts with network diversity in respect to professional background to affect innovative behaviour. This finding has straightforward implications for all organizational employees, including the managers. Our theory suggests that employees embedded in closed network of relationships will increase their innovative behaviour if they seek work-related advice from colleagues that have unlike professional background. In that way, closely embedded employees can expose themselves to diversified information and knowledge that circulates among diverse professional groups, thus offsetting the weaknesses associated with closed network position such as information and knowledge redundancy. On the other hand, employees embedded in brokering network positions will increase their innovative behaviour if they seek work-related advice from colleagues that have alike professional background. In that way, the brokers will facilitate communication and flow of resources, thus offsetting the weaknesses associated with brokering network position such as difficulty to integrate and process the available information and knowledge.

Limitations and future research

The first, and probably the most striking limitation of our study is the potential risk for endogeneity between brokerage and innovative behaviour. Indeed, the specific personal as well as contextual characteristics of individuals such as the “Big five” personality traits, expertise, talent, and experience can affect not only the innovative behaviour but also the respective position of an individual within a network of instrumental relationships (e.g. Anderson, 2008; Fang et al., 2015; Tortoriello and Krackhardt, 2010). In an attempt to minimize the potential for endogeneity, and in line with prior research in the field (e.g. Anderson, 2008; Carnabuci and Diószegi, 2015), we have

158 controlled for key personal characteristics such as intrinsic motivation and self-monitoring. We kept constant also several important contextual variables such as educational level, tenure, and hierarchical level, as context-related variables that may confound our relationship of interest (e.g. Carnabuci and Diószegi, 2015). Consistent with the practices applied elsewhere (Cheung et al., 2016; van Knippenberg et al., 2011), we ended network data collection two weeks before we approached company’s supervisors, who were asked to score the innovative behaviour of their subordinates. The different sources as well as the time lag between the two surveys should mitigate the potential risk for common method bias, and for endogeneity between our independent and dependent variables. However, we make a call for future studies that may apply longitudinal research methods to investigate further the potential for a reverse causality in the relationship between brokerage and innovative behaviour.

In order to capture the diversity of information and knowledge that circulates in an organization, in this study, we examined the moderating effect of network diversity with respect to professional background in the relationship between brokerage and innovative behaviour. Taking into consideration that the past research has shown that network diversity with respect to demographic variables such as tenure, education and gender, may affect one’s network positions (e.g. Wang et al., 2015b), future studies may investigate the moderating effect of network diversity, seen through the lenses of various categories such as age, gender, tenure, as well as personality traits.

159

REFERENCES

Adler PS and Kwon SW (2002) Social capital: prospects for a new concept. Academy of

Management Review 27(1): 17-40.

Aguinis H, Gottfredson RK and Joo H (2013) Best-practice recommendations for defining, identifying, and handling outliers. Organizational Research Methods 16(2): 270-301. Ahuja G (2000) Collaboration networks, structural holes and innovation: A longitudinal study.

Administrative Science Quarterly 45(3): 425-455.

Alessandri G and Borgogni L (2015) Stability and change of job performance across the career span. Human Performance 28(5): 381-404.

Ali M, Kulik CT and Metz I (2011) The gender diversity-performance relationship in service and manufacturing organizations. The International Journal of Human Resource Management 22(7): 1464-1485.

Amabile TM (1983) The social psychology of creativity. New York: Springer-Verlag.

Amabile TM, Hill KG, Hennessey BA and Tighe EM (1994) The work preference inventory: Assessing intrinsic and extrinsic motivational orientations. Journal of Personality and

Social Psychology 66(5): 950-967.

Ancona D and Caldwell DF (1992) Bridging the boundary: External activity and performance in organizational terms. Administrative Science Quarterly 37: 634-665.

Anderson MH (2008) Social networks and the cognitive motivation to realize network opportunities: A study of managers’ information gathering behaviors. Journal of

Organizational Behavior 29(1): 51-78.

Anderson N, Potočnik K and Zhou J (2014) Innovation and creativity in organizations: A state-of- the-science review, prospective commentary, and guiding framework. Journal of

Management 40(5): 1297-1333.

Anteby M, Chan CK and DiBenigno J (2016) Three lenses on occupations and professions in organizations. The Academy of Management Annals 10(1): 183-244.

Aral S and Van Alstyne M (2010) The diversity-bandwith tradeoff. American Journal of

Sociology. Epub ahead of print 15 April 2010. DOI: http://dx.doi.org/10.2139/ssrn.958158.

Baer M, Evans K, Oldham GR and Boasso A (2015) The social network side of individual innovation: A meta-analysis and path-analytic integration. Organizational Psychology

Review 5(3): 191-223.

Balkundi P, Kilduff M, Barsness ZI and Michael JH (2007) Demographic antecedents and performance consequences of structural holes in work teams. Journal of Organizational

Behavior 28(2): 241-260.

Bechky BA (2003) Sharing meaning across occupational communities: The transformation of knowledge on a production floor. Organization Science 14(3): 312-330.

160

Blau PM (1977) Inequality and heterogeneity: A primitive theory of social structure. New York, NY: Free Press.

Boisot MH (1995) Is your firm a creative destroyer? Competitive learning and knowledge flows in the technological strategies of firms. Research Policy 24(4): 489-506.

Boland RJ and Tenkasi JrRV (1995) Perspective making and perspective taking in communities of knowing. Organization Science 6(4): 350-372.

Borgatti SP and Foster PC (2003) The network paradigm in organizational research: A review and typology. Journal of Management 29(6): 991-1013.

Brands RA and Kilduff M (2014) Just like a woman? Effects of gender-biased perceptions of friendship network brokerage on attributions and performance. Organization Science 25(5): 1530-1548.

Burt RS (1992) Structural holes: The social structure of competition. Cambridge: Harvard University Press.

Burt RS (2000) The network structure of social capital. Research in Organizational Behavior 22: 345-423.

Burt RS (2004) Structural holes and good ideas. American Journal of Sociology 110(2): 349-399. Burt RS (2005) Brokerage and closure: An introduction to social capital. Oxford, UK: Oxford

University Press.

Burt RS (2007) Second-hand brokerage: Evidence on the importance of local structure for managers, bankers, and analysts. Academy of Management Journal 50(1): 110-145. Burt RS, Kilduff M and Tasselli S (2013) Social network analysis: Foundations and frontiers of

advantage. Annual Review of Psychology 64: 527-547.

Carlie PR (2004) Transferring, translating, and transforming: An integrative and relational approach to sharing and assessing knowledge across boundaries. Organization Science 15(5): 555-568.

Carnabuci G and Diószegi B (2015) Social networks, cognitive style and innovative performance: A contingency perspective. Academy of Management Journal 58(3): 881-905.

Černe M, Nerstad CGL, Dysvik A and Škerlavaj M (2014) What goes around comes around: Knowledge hiding, perceived motivational climate, and creativity. Academy of

Management Journal 57(1): 172-192.

Chavez-Eakle RA, Lara MC and Cruz-Fuentes C (2006) Personality: A possible bridge between creativity and psychopathology. Creativity Research Journal 18(1): 27-38.

Cheung SY, Gong Y, Wang M, Zhou L(B) and Shi J (2016) When and how does functional diversity influence team innovation? The mediating role of knowledge sharing and the moderation role of affect-based trust in a team. Human Relations 69(7): 1507-1531. Cohen WM and Levinthal DA (1990) Absorptive capacity: A new perspective on learning and

innovation. Administrative Science Quarterly 35(1): 128-152.

Coleman JS (1988) Social capital in the creation of human capital. American Journal of Sociology 94: S95-S120.

161

Cross R, Borgatti SP and Parker A (2001) Beyond answers: Dimensions of the advice network.

Social Networks 23(3): 215-235.

Cross R and Parker A (2004) The hidden power of social networks: Understanding how work

really gets done in organizations. Harvard Business School Press.

Dabbs JM, Evans MS, Hopper CH and Purvis JA (1980) Self-monitors in conservation: What do they monitor? Journal of Personality and Social Psychology 39(2): 278-284.

Damanpour F and Schneider M (2006) Phases of the adoption of innovation: Effects of environment, organization and top managers. British Journal of Management 17(3): 215- 236.

Fang R and Shaw JD (2009) Self-monitoring, status, and justice-related information flow. Journal

of Occupational and Organizational Psychology 82(2): 405-430.

Fang R, Landis B, Zhang Z, Anderson MH, Shaw JD and Kilduff M (2015) Integrating personality and social networks: A meta-analysis of personality, network position, and work outcomes in organizations. Organization Science 26(4): 1243-1260.

Fleming L and Waguespack DM (2007) Brokerage, boundary spanning, and leadership in open innovation communities. Organization Science 18(2): 165-180.

Fleming L, Mingo S and Chen D (2007) Collaborative brokerage, generative creativity, and creative success. Administrative Science Quarterly 52(3): 443-475.

Fuglestad PT and Snyder M (2010) Status and motivational foundations of self-monitoring. Social

and Personality Psychology Compass 4(11): 1031-1041.

Gilson LL, Lim HS, Luciano MM and Choi JN (2013). Unpacking the cross-level effects of tenure diversity, explicit knowledge, and knowledge sharing on individual creativity. Journal of

Occupational and Organizational Psychology 86(2): 203-222.

Granovetter M (1973) The strength of weak ties. American Journal of Sociology 78(6): 1360-1380. Grant AM and Berry JW (2011) The necessity of others is the mother of invention: Intrinsic and prosocial innovations, perspective taking, and creativity. Academy of Management Journal 54(1): 73-96.

Hammer M (1984) Explorations into the meaning of social network interview data. Social

Networks 6(4): 341-371.

Hargadon AB (1999) Group cognition and creativity in organizations. Research on managing

groups and teams 2: 137-155.

Harrison DA and Klein KJ (2007) What’s the difference? Diversity constructs as separation, variety, or disparity in organizations. Academy of Management Review 32(4): 1199-1228. Ibarra H (1992) Homophily and differential returns: Sex differences in network structure and

access in an advertising firm. Administrative Science Quarterly 37(3): 422-447.